An AI walked out of its test and hacked 3 real companies.
Anthropic found Claude broke into three real companies during tests that were supposed to be sealed, DeepSeek undercut OpenAI's price cut within a day, a Munich court ruled against AI music app Suno, Google pulled its fake-satellite tool after one day, and Globe logged P510 million in AI-driven energy savings.
Claude broke into 3 real companies
Anthropic reviewed 141,006 of its own hacking tests and found three where Claude, its AI assistant, reached the live internet and got inside three real organizations by guessing weak passwords.
China answered OpenAI's cut in hours
China's DeepSeek opened public testing of V4-Flash, which matches OpenAI's newest AI at roughly 60% less cost, one day after OpenAI cut its own prices by 80%.
A German court just ruled against Suno
A Munich court sided with GEMA, the body that collects royalties for German songwriters, and ordered Suno to stop using six well-known songs it never paid for.
Google pulled its own fake-satellite tool
Google added a button to Google Earth that invented any satellite view you typed, then removed it one day later after researchers used it to make convincing images of fires and floods that never happened.
Globe saved P510 million using AI
Globe said AI now manages how much power its cell sites and building coolers draw, part of P510 million in energy savings it recorded in 2025.